 Open Access
 Total Downloads : 1481
 Authors : V.V.K. Satyakar, Dr. J.Viswanatha Rao, S. Manikandan
 Paper ID : IJERTV1IS10423
 Volume & Issue : Volume 01, Issue 10 (December 2012)
 Published (First Online): 28122012
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Analysis of Radial Distribution System By Optimal Placement of DG Using DPSO
V.V.K. Satyakar Swarnandhra college of
Engineering & Technology, Narsapur
Dr. J.Viswanatha Rao
Swarnandhra college of Engineering & Technology, Narsapur
S. Manikandan Research Scholar,Satyabama
University,Chennai
Abstract
This paper aims to discuss the placement of Distributed Generation (DG) using Discrete Particle Swarm Optimizaiton (DPSO) algorithm in order to reduce the real power loss and improve the voltage profile. The objective function is based on real power loss reduction with relevant constraints. The proposed method deals with selection of nodes for the placement of DG and size of the DG by using DPSO. The proposed algorithm is tested with two systems consisting of 33 node and 69 node radial distribution systems.
Keywords: Distributed Generation, Radial Distribution System, Load flow, Discrete Particle Swarm Optimization

Introduction:
A distribution system is one from which the power is distributed to various users through feeders, distributors and service mains. Distribution system provides a final link between high voltage transmission systems and consumer services. The operation and planning studies of a distribution system require a steady state condition of the system for various load demands. The power losses are significantly high in distribution because of lower voltages and high currents, when compared with high voltage transmission systems. Reduction of total losses in the distribution systems is very essential to improve the overall efficiency of the distribution system. Therefore loss minimization in radial distribution systems has become the subject of intensive research. The performance of the distribution network can be extremely improved by using suitable methods like optimal placement of distributed generators.
Distributed or dispersed generation may be defined as a generating resource, other than central generating station, that is placed close to load being served, usually at
customer site. It may be connected to the supply side or demand side of meter. It can be a renewable energy source based micro, hydro, wind turbines, photovoltaic, or fossil fuel based. In terms of size, DG may range from a few kilowatts to over few megawatts. They are used to reduce power losses and to maintain a voltage profile within the acceptable limits. Proper placement of DG results in to a number of benefits like improvement of power loss reduction, improvement of voltage profile, improvement of voltage stability, systemreleased capacity, increases voltage level at the load, and improvement of voltage regulation. The extent of these benefits depends on the location, size, and number of the DGs to be installed.
The placement and sizing of generation units are important issues on an actual electrical system planning, because installing generators at certain points of the feeder can bring in one or just a few capacity gains. In some cases it is not possible to install generator only at those nodes indicated by an optimal algorithm, since independent producers and even consumers can request their access to the network in other places. In addition there may not be generation potential at some point. Even so, utilities are interested in methods which allow them to evaluate the impact of generation units in their system for reorienting them on how to deal with new producers about the possible installation points.
In the past decades, several attempts were made to improve the voltage profile and hence to reduce the power losses by placing distributed reactive power sources such as capacitor bank of optimal sizes at optimal location [1 4]. The radial distribution systems are less reliable because of its passive nature. Recently, solutions have been suggested for complementing the passiveness of RDS by embedding electrical sources of small capacity based on renewable energy technology to improve system reliability and voltage regulation [5],[6]. Such embedded
generations in the distribution system are also called as dispersed generations or distributed generations (DG).
The benefits and consequences of DG had been dealt in several papers [710]. Recently, power loss reduction has
IL (n)
S(n) *
V(n)
——— (1)
been addressed using DG, reactive power sources and network reconfiguration [11]. However, the details of reactive power management with respect to the type of DG technology employed were not discussed in [11]. This paper explores the possibility of effective power flow
where n 1, 2, 3,…..Nn
Find the Branch Current at each branch given by eqn.(2) in minor, sublateral, lateral and main lines from last node to the root node.
I (n) I (n 1) I (n 1)
control in radial distribution systems using distributed generations for improving voltage profile, voltage
b b L
where n 1, 2,3,…..Nb
——— (2)
stability index and power loss reduction in the RDS.
Particle swarm optimization (PSO) is an evolutionary computation technique developed by Kennedy and Eberhartin 1995 [12, 13]. The PSO technique has ever
Find the new voltages at each node given by eqn.(3) in main, lateral, sublateral and minor lines from root node to the end node.
since turned out to be a competitor in the field of numerical optimization. Perhaps more obvious are its ties
V(n) V(n 1) Ib (n 1) * Zb (n 1)
where n 2,3,…Nn
——— (3)
to artificial life (Alife) in general, and to bird flocking, fish schooling, and swarming theory in particular. The first version of PSO was intended to handle only nonlinear continuous optimization problems. PSO has been expanded to handle both discrete and continues variables as well.
There are various ways for obtaining the optimal location and optimal sizing of DG. In [14], sensitivity analysis is used for finding the optimal location of DG. In

the optimal location is found by finding Vindex. In

Loss sensitivity factor is used for finding the optimal location of DG. This paper finds the optimal location of DG using DPSO and also finds the optimal sizing using DPSO, compares the results with VSI and concludes that using DPSO is better option for finding the optimal location and optimal sizing of DG.


Load Flow solution of RDS:
In order to evaluate the performance of a power distribution system and to examine the effectiveness of proposed alterations to a system in the planning stage, it is essential that a load flow analysis of the system is to be carried out. One of the most fundamental and widely used analysis tool to study radial distribution system is load flow analysis.
A simple approach of backward and forward sweep [17] is carried out for load flow analysis of the given system.
Check the error between the present voltage values and previous voltage values and if the error is within limits stop the iterations, otherwise continue the same process with the obtained voltages.
Find the total Real and Reactive Power loss given by eqns.(4 & 5).
Real Power Loss is given by
n 1 b b
P Nb I (n)2 R (n) ———– (4)
Reactive Power Loss is given by
n 1 b b
Q Nb I (n)2 X (n) ———– (5)

Distributed Generation (DG):
The minimization of losses in Distribution system is very important because of the modernization of the distribution system through computer automation which requires most efficient operating scenario. In order to reduce these distribution losses DGs can be placed in the distribution system. The advantage of these DGs is reduction in real power loss, reactive power loss, improvement in the voltage profile etc.
The extent of these benefits can be achieved only through proper location and size of the DG. There are many methods of finding optimal location of DG like sensitivity analysis[14], loss sensitivity factor[16], V index[15] etc and the proposed DPSO method is compared with K. Vinoth kumar and M.P. Selvan
V1 n1
I1
R1+jX1
V2
n2
sensitivity analysis[14].
Assumptions for the placement of DG:

DG injects only real power.

The total DG capacity is limited to 30% of the total load demand.

DG location is not considered at the slack bus.
Sending End
Receiving End
P2+jQ2

Maximum numbers of DGs are limited to 3.
Fig:1 Single Line Diagram of a Branch
Find Line Current at each node given by eqn(1)
3.1 Problem Formulation:
The objective function for this algorithm aims to minimize the total power losses in the distribution system
The DPSO define each particle as a potential solution to a problem. The position of ith particle of the swarm can be represented as
along with voltage profile improvement.
Xi = (Xi1,Xi2,………. …XiN )
—– (7)
ObjectiveFunction:
min Ftotal kv fv kp fp kqfq
Nn  V(n) V(n, ref ) 
Each particle also maintains a memory of its previous best position, represented as
Pi =(Pi1,Pi2, ………PiN ) —– (8)
A particle in a swarm moves; hence, it has a velocity which can be represented as
where min fv N
n 1
n  V(n, noDG) V(n, ref ) 
Vi =(Vi1,Vi2,………….ViN ) —– (9)
min fp
n 1
P (n)
Nb withDG
n 1 Loss
P (n)
Nb withoutDG
n 1 Loss
Loss b b
(where P R I2 )
— (6)
Each swarm in the structure of swarm population is considered as MÃ—N matrix. In this matrix M represents the number of agents (population size) and N represents the number of particle is each agent.
Nb QwithDG (n)
V k 1 K V k C rand Pbest k X k
min fq
n 1 Loss Nb withoutDG
Q (n)
n 1 Loss
i i 1 1
i i — (10)
Loss b b
(where Q X I2 ) kv , kp , kq are the weighting factors.
where
C2rand2
Gbestk
Xik
System constraints:
Voltage lim its :
k Velocity of individual i at iteration k K Construction factor
V
i
C1, C2 Weight factors
X
rand1, rand2 Random numbers between [0,1]
Vmin
V(n) Vmax where n 1, 2, 3,…. Nn
k Position of individual i at iteration k
i
DG lim its :
DG DG DG 1 2 3
Pmin (n) P (n) Pmax (n) where n n DG , n DG , n DG
Pbest k Best position of individual i up to iteration k
i
Gbest k Best position of the group up to iteration k
such that PDG
PDG
PDG
PDG max
Each individual moves from the current position to the
1 2 3


Implementation of Discrete PSO for DG
next one by using the modified velocity.
Xik+1 = round Xik + Vik+1
—– (11)
Placement
4.1. Overview of the DPSO
DPSO simulates the behaviours of bird flocking. Suppose the following scenario: a group of birds are randomly searching food in an area. There is only one piece of food in the area being searched. All the birds do not know where the food is. But they know how far the food is in each iteration. So what's the best strategy to find the food? The effective one is to follow the bird, which is nearest to the food. DPSO learned from the scenario and used it to solve the optimization problems. In DPSO, each single solution is a "bird" in the search space. We call it "particle". All of particles have fitness values, which are evaluated by the fitness function to be optimized, and have velocities, which direct the flying of the particles. The particles fly through the problem space by following the current optimum particles.
The search mechanism of the DPSO using the modified velocity and position of individual based on eqns. (10) and (11) is illustrated in Figure.2.
Fig.2. The search mechanism of the DPSO

Velocity update
To modify the position of each individual, it is necessary to calculate the velocity of each individual in the next stage. In this velocity updating process, the values of parameters such as k, C1, C2 should be determined in advance.
The construction factor,
k = 2
2 – j – j2 – 4j
and j = C1 + C2
—– (12)


Results and Analysis
The effectiveness of the proposed method is
The values of C 1 and C 2 have the same value, which implies the same weights are given between Pbest and Gbest in the evolution processes.

Update of Pbest and Gbest
The Pbest of each individual i at iteration k is updated as follows
illustrated with two test systems consisting of 33node and 69node RDS. For the positioning of DG, the number of DGs is varied from one to three and the size of the DG is multiples of 50KVA. DPSO is used to find the optimal location of DG and size of the DG. The optimal location and size for the placement of DG is compared with the sensitivity analysis VSI [14] with the reduction in real power loss and the improvement in voltage profile.
Pbestk = Xk if f k > f k1
i i i i
i i
Pbestk = Pbestk1 otherwise
i i
Gbestk = best(Pbestk )
—– (13)
5.1 Example1:
The proposed algorithm is tested on 33node radial distribution system whose single line diagram is shown in
where fi, the objective function or fitness function is to be
i
evaluated at the position of individual i. Gbest at iteration
Fig.4. The variation of real power loss from one, two and three DGs are shown in Fig.5. The variation of reactive
k is set as the best evaluated position among

Stopping criteria
Pbestk .
power loss from one, two and three DGs are shown in Fig.6. The variation of voltage profile by placement of one, two and three DGs are shown in Fig.7.
The DPSO is terminated if the iteration approaches to the predefined maximum number of iterations.

Flow chart for DG Optimal Location
Start
Select parameters of DPSO:
M,N,Kmax , C1, C2 , Wmax , Wmin
Read Line & Load data of distribution system. Initialize particles with random position, no. of DGs, velocity and set k=1
Using load flow calculate the fitness value of each particle
Fig.4 33node RDS
Yes
Pbest ik
i
If k=1
No
If No
Xk
k=k+1
The summary of results is shown in Table1. The results are compared with [14] which are obtained by sensitivity analysis to identify the nodes and GA for the placement of DG. The results are shown that by placement of one
f k f (k1)
Pbest ik Pbest ik 1
i i
Yes
Pbest ik Xk i
Select Gbest at iteration 'k' among all Pbest ik
No If Boundary
conditions are satisfied
Yes
If No
k kmax
Update the Position & Velocity using Eqns (10) & (11)
Yes
Gbest is the final solution of the optimal location and size of DG
Stop
DG the real power losses reduced from 144.1389KW to 125.5972KW with improvement of 9.164% and voltage regulation has varied from 1.872% to 2.22%, by placement of two DGs the real power losses reduced from 140.8153KW to 106.1180KW with improvement of 17.15% and voltge regulation has varied from 1.88% to 4.08%, by placement of three DGs the real power losses reduced from 108.0969KW to 105.4491KW with improvement of 1.309% and voltage regulation has varied from 4.070% to 4.082%.
Fig.3 Flow chart for DG placement
Table.1 Summary results of 33node RDS with DGs
Aspect
Without DG
One DG
Two DGs
Three DGs
VSI/GA[14]
DPSO
VSI/GA[14]
DPSO
VSI/GA[14]
DPSO
Node No
–
18
12
17,18
15,32
17,18,33
14,17,32
Rating(1)KVA
–
900
1100
17800
15500
17400
14300
Rating(2)KVA
–
–
–
18100
32600
18100
17200
Rating(3)KVA
–
–
–
–
–
33600
32600
Total1100 KVA
–
900
1100
900
1100
1100
1100
Vmin (pu)
0.9131
0.9302
0.9334
0.9303
0.9504
0.9503
0.9504
Ploss (kw)
202.32
144.138
125.597
140.8153
106.1180
108.0969
105.4491
Qloss (kvar)
134.92
99.9537
83.7298
97.4054
70.1545
72.6670
69.5314
Time(sec)
–
8.6956
7.9464
8.3497
8.3754
8.5487
8.4569
Fig.5 Variation of Real power loss 33node RDS with DGs
Fig.7 Voltage Profile of 33node RDS with DGs
5.2. Example2
The proposed algorithm is tested on 69node radial distribution system whose single line diagram is shown in Fig.8. The variation of real power loss from one, two and three DGs are shown in Fig.9. The variation of voltage profile by placement of one,two and three DGs are shown in Fig.10.
36 37 38 39 40 41 42 43 44 45 46
35 36 37 38 39 40 41 42 43 44 45
51 52
50 51
66 67
65 66
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
S/S
1 2 3 4 5 6 7 8
22 23 24 25 26
9 10 11 12 13 14 15 16 17 18 19 20 21
53 54 55 56 57 58 59 60 61 62 63 64
52 53 54 55 56 57 58 59 60 61 62 63 64
68 69
67 68
Fig.6 Variation of Reactive power loss 33node RDS with DGs
65
47 48 49 50
46 47 48 49
28 29 30 31 32 33 34 35
27 28 29 30 31 32 33 34
Fig.8 69node RDS
The summary of results is shown in Table2. The results are compared with [14] which are obtained by sensitivity analysis to identify the nodes and GA for the placement of DG. The results are shown that by placement of one DG the real power losses reduced from 116.0737KW to 102.4029KW with improvement of 6.077% and voltage regulation has varied from 5.125% to 4.855%, by placement of two DGs the real power losses reduced from 107.0353KW to 101.7987KW with improvement of 2.33% and voltage regulation has varied from 5.165% to 5.063%, by placement of three DGs the real power losses reduced from 103.0896KW to 101.7719KW with improvement of 0.585% and voltage regulation has varied from 5.182% to 5.101%.
Fig.10 Voltage Profile of 69node RDS with DGs
Fig.9 Variation of Real power loss 69node RDS with DGs
Table.2 Summary results of 69node RDS with DGs
Aspect
Without DG
One DG
Two DGs
Three DGs
VSI/GA[14]
DPSO
VSI/GA[14]
DPSO
VSI/GA[14]
DPSO
Node No
–
65
61
64,65
61,64
63,64,65
61,64,65
Rating(1)KVA
–
651150
611150
641100
61850
63850
61850
Rating(2)KVA
–
–
–
6550
64300
64250
64250
Rating(3)KVA
–
–
–
–
–
6550
6550
Total1150 KVA
–
1150
1150
1150
1150
1150
1150
Vmin(pu)
0.9091
0.9557
0.9533
0.9561
0.9552
0.9563
0.9555
Ploss(kw)
224.95
116.073
102.40
107.0353
101.79
103.08
101.77
Qloss(kvar)
102.1472
56.4442
49.5794
51.9056
49.2760
49.9240
49.2625
Time(sec)
–
27.0619
25.7261
26.3369
26.3802
27.1227
27.7273


Conclusion
DPSO for solving the DG placement in RDS has been proposed in this paper. This paper aims at discussing the reduction of real power loss and maintenance of voltage profile by using DG . The proposed method deals with
optimal selection of nodes for the placement of DG and Size of the DG by using Discrete Particle Swarm Optimization (DPSO). From results the reduction of Real power loss and the improvement of voltage profile can be observed. Also that for the same capacity of DGs for
single, two and three DGs the results are better if the number of DGs are more, especially the minimum voltage is improved if we have more number of DGs being total capacity same. The proposed algorithm is tested with two systems consisting of 33 node, and 69 node RDS. From the results, several important observations can be concluded as follows.
From the results, important observations can be concluded as follows.

The power losses of distribution system can be effectively reduced by proper placement of DG.

In addition of power loss reduction, the voltage profile can be improved.
References:

M.E. Baran, and F.F. Wu, Optimal capacitor placement on radial distribution system, IEEE Trans. Power Delivery, Vol. 4 (1), pp. 725 734, 1989.

M.E Baran, and F.F. Wu, Optimal sizing of capacitors placed on radial distribution system, IEEE Trans. Power Delivery., Vol. 4 (1), pp. 735 743, 1989.

Srinivasan sundhararajan and Anil pahwa, Optimal selection of capacitors for radial distribution systems using a genetic algorithm, IEEE Trans Power System, Vol. 9 (3), pp. 1499 1507, 1994.

Y.C.Huang, H.T.Yang,, C.L.Huang, Solving the capacitor placement problem in a radial distribution system using Tabu search approach,IEEE Trans. ower System, Vol. 11 (4), pp. 1868 1873, 1996.

T. Ackermann, G. Anderson, and L.S. Soder, Distributed generation: a definition, Electric Power Systems Research, Vol. 57 (3), pp. 195 204, 2001.

N. Jenkins, R. Allan, P. Crossley, D. Kirschen, and G.Strbac, Embedded generation, The institution of electrical engineers, London, 2000, 1st edition

G. Carpinelli, G. Celli, F. Pilo, A. Russo, Distributed generation siting and sizing under uncertainity, IEEE Porto power tech proceedings, Porto, Portugal, Sept. 2001, 4, pp.1 7

A. Bhowmik, A. Maitra, S.M. Halpin, E. Schatz, Joe, Determination of allowable penetration levels of distributed generation resources based on harmonic limit considerations, IEEE Trans. Power Delivery, Vol. 18 (2), pp. 619 624, 2003.

T. Niknam, A.M. Ranjbar, A.R. Shirani, Impact of distributed generation on Volt/Var control in distribution networks, IEEE Bologna power tech conference, Bologna, Italy, June 2326, 2003, pp.1 7.

G.W. Ault, J.R. McDonald, and G.M. Burt, Strategic analysis framework for evaluation distributed generation and utility strategies, IEE Proc. Generation, Transmission, Distribution, Vol. 150 (4), pp.475 481, 2003.

M.E. Hamedani Golshan, and S.A. Arefifar, Distributed generation, reactive sources and network configuration planning for power and energyloss reduction, IEE Proc. Generation, Transmission,Distribution, Vol.153 (2), pp.127 136, 2006.

J. Kennedy and R. Eberhart, "Particle Swarm Optimization", Proceedings of IEEE International Conference on Neural Networks (ICNN'95), Vol. IV, pp.19421948, Perth, Australia, 1995.

J. Kennedy ,R. Eberhart, A new optimizer using particle swarmtheory, Proc. 6th Int. Symposium on Micro Machine and Human Science, pp. 3943, 1995.

K. Vinoth kumar and M.P. Selvan , Planning and Operation of Distributed Generations in Distribution Systems for Improved Voltage Profile , 2009 IEEE , pp.1 7, 2009.

J. B. V. Subrahmanyam, C. Radhakrishna Distributed Generator Placement and Sizing in Unbalanced Radial Distribution System , International Journal of Electrical and Electronics Engineering, Vol.3, pp.746 753. 2009

Satish Kansal1, B.B.R. Sai, Barjeev Tyagi, and Vishal Kumar , Optimal placement of distributed generation in distribution networks , International Journal of Engineering, Science and Technology , Vol. 3, No. 3, 2011, pp. 4755

K. Vinoth kumar and M.P. Selvan , A Simplified Approach for Load Flow Analysis of Radial Distribution Network , International Journal of Computer and Information Engineering Vol 2 No 4 ,2008,pp 272283

M.Chakravorty, D.Das, Voltage Stability analysis of radial distribution networks, International Journal of Electrical Power and Energy Systems, Vol.23(2),pp129135,2001